Osteoarthritis (OA) is the most common degenerative joint disease, afflicting ~30 million US adults and currently lacking effective treatment. OA diagnosis with radiography and MRI relies predominantly on biomarkers of late disease, and capability for early detection of OA could enable therapies preventing irreversible cartilage loss. Changes in subchondral bone (SB) architecture and bone mineral density (BMD) have been identified as possible biomarkers of early disease and a target for therapy. However, this potential advance in OA diagnosis and treatment is hampered by the lack of imaging technology that combines the high level of spatial resolution required to characterize trabecular architecture (<0.2 mm) with a high degree of quantitative accuracy enabling reliable BMD measurement and the ability to image large joints in vivo. Dedicated extremity cone-beam CT (CBCT) is a recent advance providing the capability for weight-bearing imaging, reduced dose, and superior spatial resolution (~0.3 mm) compared to conventional CT, as demonstrated in clinical studies using a system developed at our institution. We hypothesize that an advanced system for extremity CBCT utilizing a novel CMOS detector, model-based image reconstruction, and patient-specific scatter correction will provide simultaneous in-vivo assessment of SB architecture, BMD, and joint space morphology for early diagnosis and evaluation of treatment response in OA. The development and clinical translation of the system supporting this hypothesis involve the following Specific Aims: 1) Develop an advanced extremity CBCT configuration for in-vivo assessment of bone health employing a fast, low-noise, high-resolution CMOS detector within a system optimized for OA imaging tasks and compatible with rapid translation to the extremity CBCT clinical prototype; 2) Enhance spatial resolution in CBCT beyond conventional limits with model-based reconstruction incorporating a model of system blur, fiducially-free rigid motion correction, and a volume-of-interest, multi-resolution object representation to yield minimum resolvable detail size of ~0.1 mm to facilitate accurate SB morphometry; 3) Improve quantitative accuracy and soft tissue contrast (e.g., cartilage and tendons) using a patient-specific scatter correction with Monte Carlo calculation accelerated with iterative de-noising (anticipated correction time ~2 min/scan) and by developing a novel, automatic BMD calibration system integrated with the scanner enclosure to yield BMD accuracy better than 3%; and 4) Integrate the systems from Aims 1-3 onto a clinical platform for extremity CBCT and evaluate the system in a clinical pilot study to assess potential diagnostic performance and reproducibility of the SB architecture and BMD measurement. The project will advance the resolution and quantitative accuracy of CBCT translated to pilot studies pertinent to the early detection and staging of OA, with potential application in other areas of bone health e.g. rheumatoid arthritis or osteoporosis. The research is conducted in an academic-industry consortium between Johns Hopkins University and Carestream Health.